You are on page 1of 11

Applied Clay Science 42 (2009) 368378

Contents lists available at ScienceDirect

Applied Clay Science


j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / c l a y

X-ray diffraction studies of the thermal behaviour of commercial vermiculites


C. Marcos a,, Y.C. Arango b, I. Rodriguez a
a b

Dpto. Geologa & Inst. de Qumica Organometlica Enrique Moles, Univ. Oviedo, Jess Arias de Velasco s/n, 33005, Oviedo, Asturias, Spain Laboratorio de Fsica del Plasma, Universidad Nacional de Colombia, Sede Manizales, A.A 127, Manizales (Caldas), Colombia

A R T I C L E

I N F O

A B S T R A C T
For the purpose of knowing the vermiculites which would have larger capability to retain contaminating substances heating commercial samples from different places have been identied and their thermal behaviour at several temperatures has been investigated by X-ray diffraction (XRD), electron microprobe, thermal analysis (TG and DTA), scanning electron microscopy (SEM), and transmission electron microscopy (TEM). The commercial vermiculites can be divided into two types: type 1 (Sta. Olalla, Piau and Gois) with Mg2+ and K+ (b 1) as the principal cations in the interlayer space, and type 2 (China E, China W, China G and Palabora) with K+ (approximately = or N 1) and/or Na+ and/or Ca2+ with or without Mg2+ as the principal interlayer cations. The process of dehydration in situ with the temperature seems restricted to interlamellar water monolayers of 1WLHS type-1 of vermiculites, without dehydration to a zero-water-layer-hydration state (0-WLHS) and the dehydroxylation starts at lower temperatures than in vermiculites of type 2. The maximum hydration state exhibited by the type-2 samples at ambient temperature was equal or lower than the monolayer hydrate, the dehydration process in situ with the temperature was slower and the dehydrated vermiculite coexists with a mica-like structure. The behaviour of vermiculites at elevated temperature examined in situ can be understood considering that the vermiculites constitute a complex system not necessarily in equilibrium and where kinetics plays an important role. Commercial vermiculites heated abruptly at 1000 C during 1 min transform to mica-like or mica-like coexisting with enstatite, in contrast to the purest Sta. Olalla vermiculite, with only magnesium interlayer cations, which changes to enstatite. 2008 Elsevier B.V. All rights reserved.

Article history: Received 8 August 2007 Received in revised form 4 March 2008 Accepted 8 March 2008 Available online 14 March 2008 Keywords: Electron microscopy Thermal analysis Vermiculite X-ray diffraction

1. Introduction Vermiculite found important industrial applications (Strand and Stewart, 1983; Hindman, 1992; Bergaya et al., 2006). Vermiculite is of secondary origin and results from alteration of macroscopic particles of biotite and iron-bearing phlogopite, chlorite, pyroxene, or other similar minerals by weathering, hydrothermal action, percolating groundwater, or a combination of the three effects (Basset, 1963). In natural form, it has the size and shape of mica, presents disorder effects, and the ability of dehydrationrehydration and swelling processes (Mathieson and Walker, 1954; Shirozu and Bailey, 1966; Grim, 1968; Brown and Brindley, 1980; de la Calle and Suquet, 1988). Commercial vermiculite, which is considered here, is normally a beneciated product composed of particles larger than 1 mm in size. Vermiculites have water interlamellar layers and, consequently, can undergo processes of dehydrationhydration which depend on temperature, pressure, chemical composition, particle size and relative humidity (Mathieson and Walker, 1954; Vali and Hesse, 1992; Collins et al., 1992; Reichenbach and Beyer, 1994, 1995; RuizConde et al., 1996; Marcos et al., 2003).
Corresponding author. Tel.: +34 985 10 31 00; fax: +34 985 10 31 03. E-mail address: cmarcos@uniovi.es (C. Marcos). 0169-1317/$ see front matter 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.clay.2008.03.004

The hydration properties of 2:1-clay minerals are controlled to a large extent by the interlayer cation (for instance, see Hendricks et al., 1940; Mooney et al., 1952; Suquet et al., 1975; Bergaya et al., 2006). Divalent cation vermiculites were found to contain up to a maximum of two water layers. Swelling was absent in K+-vermiculite, although most K+-vermiculites retain a basal spacing of about 10.4 at all relative humidities. Contraction by K+-saturation may be absent or only partial with low-charge vermiculites. Here a distinct 12.5 spacing is observed and in many vermiculites a broad intermediate spacing indicates interstratication between contracting and non-contracting forms (Walker, 1961). The inuence of the interlayer region on the structure of the silicate layer is also reected to some extent by the effect of the interlayer cations on the length of the b axis. The variation of the b dimension of the laboratory-produced dioctahedral vermiculite varied with the interlayer cation and with its hydration state. It is assumed that (Leonard and Weed, 1970) dehydration involves rotation of the surface oxygen triads of the silicate layer, whereas in the hydrated state the Mg ions stabilize a water network with sufcient energy to limit such rotation. The hydration state of vermiculite is dened by the number of water layers in the interlamellar space, such as zero-, one- and twowater layer hydration states (0-, 1- and 2-WLHS, respectively, notation after Suzuki et al. (1987) and used in the present paper).

C. Marcos et al. / Applied Clay Science 42 (2009) 368378

369

Fig. 1. XRD patterns of commercial vermiculites at ambient temperature.

370

C. Marcos et al. / Applied Clay Science 42 (2009) 368378 Table 3 Cationic distribution in the structure (tetrahedral, T; octahedral, O sheet and interlayer spaces, I) and Si/AlIV ratio Sta. Olalla Si4+ Al3+ Fe(total) T Al3+ Ti4+ Fe(total) Cr3+ Mn2+ Mg2+ O Mg2+ Ca2+ Na+ K+ inter Si/AlIV 5.66 2.34 0.00 8.00 0.59 0.04 0.43 0.00 0.02 4.92 6.00 0.75 0.05 0.04 0.01 0.85 2.43 Piau 6.09 1.66 0.24 8.00 0.00 0.13 0.61 0.01 0.01 5.24 6.00 0.55 0.03 0.01 0.69 1.28 3.67 Gois 6.46 1.54 0.00 8.00 0.61 0.09 1.27 0.00 0.01 4.01 5.99 0.25 0.01 0.04 0.20 0.50 4.20 China W 4.88 1.71 0.46 7.04 0.00 0.09 0.00 0.02 0.00 4.43 4.54 0.00 0.05 0.17 1.17 1.39 2.86 China G 6.22 1.78 0.00 8.00 0.48 0.15 1.16 0.05 0.01 4.16 6.00 1.52 0.18 1.26 1.21 4.17 3.49 Palabora 6.70 1.30 0.00 8.00 0.61 0.15 1.07 0.00 0.00 4.16 6.00 1.50 0.03 0.05 1.25 2.83 5.17 China E 4.60 2.06 1.20 7.87 0.00 0.11 0.37 0.00 0.01 2.75 3.25 0.00 0.15 0.14 0.99 1.29 2.23

Table 1 Hydration states (0/1- and 2-water-layer-hydration (WLHS)) and interstratied phases (I) Basal spacing ( ) 2-WLHS Sta. Olalla Piau Gois China W China G Palabora China E 14.5 14.5 14.7 14.5 13.2 11.9 10.2 11.9 23.2 12.6 25.2 12.4 12.3 12.3 1-WLHS 1/0-WLHS I
V

When vermiculite is rapidly heated to high temperatures, the evaporating water molecules increase the particle volumes. This process of thermal expansion produces a lightweight product which nds use in various construction products, agriculture, horticulture, and other applications (Strand and Stewart, 1983; Hindman, 1992; Bergaya et al., 2006).
2. Experimental methods: samples and techniques The vermiculite samples investigated were given by suppliers. They come from Catalo (Gois, Brasil), Paulistana (Piau, Brasil), Palabora (South Africa) and three from China (named East, West and G), respectively. Vermiculites from Sta. Olalla (Huelva, Spain) was also collected for comparison. The Gois vermiculite is associated to an ultra-maphic complex. The vermiculite from Piau occurs in a hybrid basic rock, probably a lamprophyre (Hennies and Stellin Junior, 1978). In the Palabora Complex at Republic of South Africa, phlogopite is progressively transformed into vermiculite following the sequence: phlogopite mixed-layer vermiculite (Badreddine, 1988). Although the alteration of phlogopite into vermiculite is considered to be the result of a weathering process caused by ordinary groundwater, Gevers (1949) proposed the inuence of magmatic uids to explain the origin of vermiculite. The present authors do not know the origin of the China vermiculites and available references were not found. The origin and mineralogy of the Sta. Olalla vermiculite has been studied extensively (Gonzlez Garca and Garca Ramos, 1960; Velasco et al., 1981; Justo, 1984; Luque et al., 1985). This vermiculite is formed from a mica identied as a phlogopite resulting from the alteration of pyroxenites. Two experiments using a Seifert XRD 3000 diffractometer (Servicios Cientco-Tcnicos de la Universidad de Oviedo) have been carried out: 1) using samples as akes at room temperature, for the purpose of vermiculite characterization; 2) after heating the sample as powder to 1000 C during 1 min in an oven once stabilized (~2 h), for the purpose of knowing the treated product. The machine settings were 30 mA and 40 kV (Cu-K radiation; =1.5418 ), 2 range 320, 2 step scans of 0.02 and a counting time of 20 s per step. On the other hand, it has been used a Bruker AXS diffractometer (Laboratorio de Fsica del Plasma, Universidad Nacional, sede Manizales) at different temperatures in situ, in the range between 20 C or 30 C and 1000 C, collected each 10 C in the range of 20 C to 200 C and each 100 C in the range of 200 C to 1000 C, in order to study the dehydration process and the resultant products in the ake samples. The machine settings were 30 mA and 40 kV (Cu-K radiation; = 1.5418 ), 2 range 340, 2 step scans of 0.1 and a counting time of 20 s per step. The temperature of the sample was measured by a thermocouple. A platinum band connecting the two heating electrodes acted as a sample holder and also as a heat source. The temperature registered by the thermocouple closely resembles the sample temperature because the samples form a thin layer on the surface of the platinum band. The experiments with Bruker AXS diffractometer were performed without vacuum because Marcos et al. (2003) demonstrated that dehydration in the vacuum is similar to that by heating. When both temperature and vacuum are acting, the sample dehydrates just after the vacuum is established and the temperature has no further effect. Table 2 Chemical analyses (wt.%) Sta. Olalla SiO2 TiO2 Al2O3 Cr2O3 FeO MnO MgO CaO Na2O NiO K2O Total 35.93 0.33 15.78 0.03 3.27 0.14 24.13 0.29 0.12 0.00 0.03 80.05 Piau 39.94 1.12 9.26 0.06 6.69 0.04 25.47 0.20 0.04 0.02 3.54 86.41 Gois 40.67 0.79 11.51 0.01 9.58 0.08 18.05 0.03 0.12 0.01 1.07 81.94 China W 43.22 1.01 11.87 0.16 4.28 0.01 24.27 0.40 0.71 0.04 7.48 93.46 China G 35.65 1.16 11.00 0.39 4.63 0.04 21.81 0.92 3.55 0.06 5.61 84.80 Palabora 41.07 1.25 9.96 0.02 7.85 0.03 23.30 0.18 0.15 0.06 6.01 89.88 China E 36.61 1.16 13.91 0.03 14.92 0.12 14.75 1.17 0.60 0.06 6.49 89.82

Chemical analyses of the samples were performed using a CAMEBAX-MBXSX-50 electron microprobe (Servicios Cientco-Tcnicos de la Universidad de Oviedo), using an acceleration voltage of 15 kV and a beam current of 15 nA. Analyses on 5 different akes (1 to 5 mm2) were performed in each sample. The cations were obtained using the MINPET program (MINPET GEOLOGICAL SOFTWARE, 146 DV Chateau MassonAngers, Quebec, Canada).The calculations to obtain the cations are based on 24 anions (20 structural oxygen and 4 OH groups). The thermogravimetric (TG) analyses were made with a Mettler Toledo Stare System. The heating rate was 10 C/min. The total weight loss was determined gravimetrically by heating the samples at 1000 C in a mufe furnace and assumed due entirely as water. The scanning electron microscopy (SEM) using a JEOL-6100 equipment (Servicios Cientco-Tcnicos de la Universidad de Oviedo), was used to take images of exfoliated vermiculites after heating abruptly at 1000 C during 1 min. For high-resolution transmission electron microscopy (HRTEM) the samples were gently pulverized not to introduce changes in the crystallinity and then deposited on

Fig. 2. X-ray diffraction patterns in situ of Sta. Olalla sample.

C. Marcos et al. / Applied Clay Science 42 (2009) 368378

371

the value 100 due to the limitation of electron microprobe analysis of light elements, such as hydrogen and oxygen. The percentage of the element as oxides agrees with those previously published (Fster, 1963; Grim, 1968; Justo, 1984; Justo et al., 1986). The possible differences among them may be explained taking into account the presence of other elements in minor percentage, such as Cr2O3, MnO, etc. (Fster, 1963; Norris, 1973). The K2O content b 0.35 suggests that Sta. Olalla is pure vermiculite (Justo et al., 1986). The other samples analyzed are not pure vermiculites because the percentage of K2O is higher. This is in agreement with the X-ray results. The distribution of cations in the structure calculated according to Fster (1963) is presented in Table 3. Sta. Olalla, Gois, China G and Palabora samples have aluminium and iron ions as the main isomorphous substitution for octahedral magnesium ions. Piau and China E samples have iron ions replacing magnesium ions. The content of tetrahedral and octahedral cations is 100% except in China W with values of 88 and 75.7%, respectively, and China E with 98.4 and 54.2%, respectively. The interlayer Mg2+-cation content was zero in these last samples. Due to the interlayer cations, the commercial samples investigated can be divided in two types: type-1 (Piau and Gois) with Mg2+ or Mg2+ and K+ (b 1) as the principal interlayer cations in the interlayer space, and type-2 (China E, China W, China G and Palabora) with K+ (approximately = or N 1) and/or Na+ and/or Ca2+ with or without Mg2+ as the principal cations in the interlayer. Vermiculite of Sta. Olalla corresponds to type 1. 3.2. X-ray diffraction experiments in situ 3.2.1. Sta. Olalla The XRD diagrams for Sta. Olalla sample (Fig. 2) from 20 to 1000 C showed a progressive displacement of reections and changes in their. Thus structural changes were taking place at increasing temperature.

Fig. 3. X-ray diffraction patterns in situ of Piau sample.

copper grids for TEM. The observations were made using a JEOL-2000 EX-II instrument (Servicios Cientco-Tcnicos de la Universidad de Oviedo). All images were taken in bright eld illumination using an accelerating voltage of 200 kV.

3. Results and discussion 3.1. Identication The X-ray patterns obtained at room temperature for the samples are presented in Fig. 1. The hydration states (1/0-1-2-water-layerhydration (WLHS)) and interstratied phases are shown in Table 1. Both Piau and Sta. Olalla samples showed the (002) reection, at 6.2 2, which is the most characteristic reection of vermiculite (JCPDS card 16-613). The diffraction pattern of Gois sample also shows the (002) reection as the more intense reection although it is accompanied by further broadened reections (25.2 , 12.4 ) with lower intensity and more broader, clearly indicating that the crystallinity of this material is much poorer than that observed in Sta. Olalla and Piau samples. They represent interstratied phases. In the X-ray patterns of China W, China G, Palabora and China E samples the intensity of the reections is very low, indicating the poor crystallinity of these materials as in Gois sample. The X-ray pattern of China W indicated the coexistence of different states of hydration 2WLHS (14.5 and 13.5 ) with the 1-WLHS (11.3 ) and other interstratied phases (12.8 , 12.0 ). The X-ray pattern of China G revealed the coexistence of an interstratied phase with d-spacing of 12.4 and a dehydrated phase 1/0-WLHS (10.2 ). In Palabora sample the more intense reection corresponds to a dehydrated phase, 1WLHS (11.9 ) which has a shoulder at 12.6 . The more intense reection of China E sample corresponds to a dehydrated phase, 1/0WLHS, with basal spacing of 10.4 . The chemical analysis data of all the vermiculite samples studied are listed in Table 2. The sum of the percentage of elements does not reach

Fig. 4. X-ray diffraction patterns in situ of Gois sample.

372

C. Marcos et al. / Applied Clay Science 42 (2009) 368378

The diagrams of Fig. 2a show different types of variation in the displacement of the diffraction lines and/or intensity proles. The rst characterizes the transition at 90 C between 2-WLHS phases (14.5 to 13.8 ) with a different arrangement from that in the 14.5 phase. It consists of progressive displacement of the diffraction lines with slight changes in their maximum intensities. The 13.8 phase persists only up to 100 C. The second type of variation occurs in the dehydration to 1-WLHS (11.6 ) at 100 C, where this phase coexists with 13.8 and 12.8 phases. At 140 C the 11.5 phase has a more symmetrical prole, with increased intensity. The third type of variation leads to the 1-WLHS 11.5 and 10.8 phases, at 300 C, with higher reection intensities. The fourth variation provokes a sudden decrease of intensities at 400 C, where a new reection appears at 10 . Its intensity increases until 800 C, and disappears at 1000 C. The coexistence of hydrated states with different d-values during the dehydration process has also been reported for other Mg-vermiculites (Collins et al., 1992; Reichenbach and Beyer, 1994; Ruiz-Conde et al., 1996; Marcos et al., 2003). Under these conditions, the 0-WLHS is never achieved. The 12.8 phase is an interstratied phase according to previous studies (Collins et al.,1992; Ruiz-Conde et al.,1996; Marcos et al., 2003) and it is probably originated from two- and one-sheet hydrates. On the other hand, in the 2 range 2040 (Fig. 2b) it is observed that the 008 (3.62 ), 0,0,10 (2.88 ) and 0,0,12 (2.40 ) reections are present up to 100 C although they show shoulders at 80 C (Fig. 2b).

Fig. 6. X-ray diffraction patterns in situ of China G sample.

Fig. 5. X-ray diffraction patterns in situ of China W sample.

Fig. 7. X-ray diffraction patterns in situ of Palabora sample.

C. Marcos et al. / Applied Clay Science 42 (2009) 368378

373

interstratication with a d-value of 12.8 . The aforementioned 1WLHS phase remains up to 300 C at which the phase with a d-spacing of 10.6 appears. Above this temperature, this reection shifts to higher values of 2 and its intensity diminishes very much. In this case, a greater dehydration than in Sta. Olalla vermiculite is achieved, as it occurs in vacuum (Marcos et al., 2003), although the presence of the 0WLHS phase (about 9 ) is not very clearly indicated. In the range of 2040 of 2 (Fig. 3b), all reections disappear above 300 C. 3.2.3. Gois The XRD patterns of Gois akes (Fig. 4) indicated the coexistence of the (14.7 ) 2-WLHS phase with the interstratied phases with d = 12.4 and 25.2 up to 80 C. At 90 C the (13.8 ) 2-WLHS phase coexists with the 12.1 and 25.2 interstratied phases. At 100 C the (11 ) 1/0-WLHS phase coexists with the 11.9 and 22.4 interstratied phases and the intensity of reections begins to increase up to 200 C. At 300 C the (10.5 ) 1/0-WLHS phase is only the single phase present and the intensity decreases again. At 500 C this phase disappears practically and new reections are observed, being more intense than that one at 2.46 (Fig. 4b), which corresponds to forsterite (JCPDS card 34-189). From 600 to 700 C forsterite coexists with mica. From 700 C forsterite begins to disappear and at 1000 C mica is practically the unique phase present. 3.2.4. China W The diffraction patterns of China W (Fig. 5a) at 30 C show the coexistence of 2-WLHS phases 14.5 and 13.5 with other phases at 12.8 , 12.0 and 11.3 ) (14.5 and 13.5 ) with the 1-WLHS (11.3 ) and other interstratied phases (12.8 , 12.0 ). At 40 C there is a structural collapse and the reection had a d-spacing of 12.0 with two shoulders. At 80 C, the reections were broadened and of lower intensity. From 90 to 110 C we observe very broadened reections at 12.8 , 12 , 11.2 and 10.510.2 . From 120 C to 300 C the intensity of the reections increases and at 140 C the 10.3 reection was pronounced. Above 300 C, the structure collapsed and then the 10.3 peak disappears and the reections with d-spacing of 12.8 , 11.8 and 11.2 coexists and the reection at 9.99.7 (0WLHS phase) begins to appear. In general, small crystallites of mica, probably muscovite-like (according to JCPDS card 7-25), coexisted with vermiculite with different hydration states. 3.2.5. China G The reection with d-spacing of 12.4 shifts gradually to 10.2 at 300 C, and its intensity remains almost constant above 60 C. Starting at 600 C, the intensity of the reection at 10.2 decreases abruptly

Fig. 8. X-ray diffraction patterns in situ of China E sample.

At 110 C the reection (0,0,12) changes into a band. At this temperature the intensity reaches its lowest value and increased again above 140 C where the pattern is similar to that at 40 C. At higher temperatures the intensity of the reections decrease until 400 C at which the structure a collapse. At 500 C the reection with a d-spacing of 3.303.35 appears and its intensity increases until 800 C, before it decreased again. 3.2.2. Piau In the XRD diagrams of akes of Piau (Fig. 3a) the (14.5 ) 2-WLHS phase transforms to 2-WLHS phase with d-value of 13.8 at 70 C. Then it transforms into the (11.6 ) 1-WLHS phase through the

Fig. 9. Schematic thermal transformation as indicated registered by XRD (in situ).

374

C. Marcos et al. / Applied Clay Science 42 (2009) 368378

Fig. 10. TG and DTA curves related to changes in basal spacings during heating. a) Sta. Olalla, b) Piau, c) Gois, d) China W, e) China G, f) Palabora, g) China E.

indicating the transformation into a random interstratication of 1and 0-WLHS phases. It may also indicate a second-order reection of regular interstratication 1- and 0-WLHS (Collins et al., 1992). The

10.2-A spacing could also be interpreted as muscovite (JCPDS), because in the 2 range of 2030 (Fig. 6b), the more intense reection at 6001000 C occurs at 3.3 A, typical of muscovite.

C. Marcos et al. / Applied Clay Science 42 (2009) 368378

375

3.2.6. Palabora The XRD patterns of Palabora sample (Fig. 7) showed variations both in the intensity and at the position of reections, without abrupt changes. The reection at 20 C with a d-spacing of 11.9 , and a shoulder at 12.4 , was shifted to a value of 11.3 at 90 C and its intensity decreased. This reection shifted to about 11 up to 300 C, rstly with increasing intensity. At 300 C the structure collapsed, although a broad reection at 10.9 , with a shoulder at 10.5 was seen. Above 300 C, the reection with a d-spacing of 10.510.2 and the appearance of the reections at 3.3 , 2.5 and 2.3 (Fig. 7b) indicated the formation of a new micaceous crystalline phase, probably biotite (JCPDS card 24-967). 3.2.7. China E The (002) reection maintains during the collection of data with a d-spacing of 1010.2 . This reection together with those ones with a d-spacing about of 3.56, 3.15, 3.0, 2.90, 2.50 correspond to dehydrated vermiculite (Fig. 8b). Above 300 C these last reections disappear and changes are observed in the intensity of the (002) reection. This change may be attributed to formation of a mica-like phase and to dehydrationrehydration of the sample as well. In contrast, this reection has been interpreted (Collins et al., 1992) as a second-order reection arising from and ordered alternation of oneand zero-layer hydrates, but could also correspond to a random interstratication of one- and zero-layer hydrates. 3.2.8. Summary of in situ XRD studies The samples at ambient temperature have 2-WLHS (Sta. Olalla, Piau, Gois and China W) and dehydrate in dened states. Dehydration seems to be inhibited at a one-water layer hydration state (1-WLHS), without dehydration into a zero-water layer hydration state (0-WLHS), as it occurs under vacuum (Marcos et al., 2003). The dehydration of these samples is more signicant as compared to those with lower potassium content due to the relatively low hydration energy of the K+ ions.

Table 4 Water contents (%) and molar ratio H2O/interlayer cation derived from TG T C Sta. Olalla 20 100 140 300 20 90 300 20 100 200 300 30 40 50 90 140 400 30 40 50 70 100 120 300 20 90 100 300 20 50 300 d002 14.5 11.6 11.5 10.8 14.5 11.6 10.6 14.7 11.9 10.5 10.3 14.5 12.6 11.9 11.3 10.3 9.9 13.0 11.8 11.5 11.3 10.6 10.5 10.4 11.9 11.3 11.0 10.5 10.0 10.2 10.2 mol H2O/mol interlayer cations 13.32 9.24 7.01 3.73 10.4 8.4 3.1 23.62 15.97 9.23 7.07 7.5 7.3 7.0 5.6 4.4 3.7 1.5 1.5 1.4 1.2 1.1 0.9 0.8 4.9 3.8 3.7 2.3 7.9 7.5 2.7 Water content % 25.6 17. 8 13.5 7.2 14.0 11.0 5.2 13.6 9.19 5.3 4.1 17.6 17.1 16. 6 13.3 10.4 5.3 12.3 12.2 11.7 9. 8 7.0 7.7 6.4 13.4 10.6 10.2 6.4 13.1 12.3 5.3

Piau

Gois

China W

China G

Palabora

Fig. 11. Ex situ XRD patterns of commercial vermiculites abruptly heated to 1000 C (Symbols meaning: E = Enstatite, M = Mica).

China E

376

C. Marcos et al. / Applied Clay Science 42 (2009) 368378

The maximum hydration state of the other samples (Palabora, China E and China G) at ambient temperature was equal or lower than a one-layer hydrate and dehydration was slower. In last stage, thermal dehydration of vermiculites takes place by dissociation of protons from OH groups, followed by migration to neighbouring OH positions where they react to form H2O, which is then released from the structure by diffusion. The migration of water depends on the interlayer cations: cations of large ionic radii should cause greater hindrance than smaller ones (Serratosa and RausellColom, 1985). Therefore, dehydroxylation is hindered by the presence of interlayer cations trapped within the ditrigonal cavities in adjacent layers. Fig. 9 illustrates dehydration, dehydroxilation and transformations. Dehydroxylation begins at lower temperature for samples with lower content of potassium ions and with 2-WLHS at ambient temperature. In contrast to Sta. Olalla and Gois vermiculites, the dehydrated vermiculites of Piau, Palabora, China E, China W and China G in which the content of interlayer cations of large ionic radii is higher than of the smaller cations, coexist with a mica-like structure. 3.2.9. Thermal analysis (TG and DTA) The changes in the basal spacing during heating of the vermiculites (Fig.10) are in agreement with the X-ray diffraction studies. For the Gois sample three regions on the TG and XRD curves can be distinguished and for China W two regions. In the other cases, three and two regions on the TG and XRD curves can be distinguished. The third region in the XRD pattern curve of Palabora could not be identied because a higher resolution is probably needed. No region can be distinguished in XRD curve of China E sample. The different numbers of the regions can be explained by differences between both techniques (Reichenbach and Beyer, 1995) and in the time and kinetic effects. The transitions between them are marked by endothermic peaks on the DTA curve and they occur between room temperature and 300 C. Thus, the total weight loss assigned as water content can be taken as a basis for calculating the number of water molecules of the interlayer cations (Table 4). The results are consistent with the scheme proposed by Serratosa and Rausell-Colom (1985): The molar ratio H2O/interlayer ion is higher in samples with lower content of cations in the interlayer space. In the 14.5 phase of Sta. Olalla vermiculite, in which Mg2+ is practically the unique interlayer ion, this ratio closely resembles the value given by Walker and Cole (1957) who used the same heating rate (10 C/min), and those by Alcover and Gatineau (1980); the ratio obtained for the 11.6 and

Fig. 13. SEM image of isolated plates of minerals on the surface of the pyroexpanded sample of China E (as an example).

11.5 phases agrees with that given by Walker and Cole (1957). Values obtained for Piau and Gois samples which have also K+ ions in the interlayer space are close to the Sta. Olalla values. 3.2.10. Ex situ X-ray diffraction experiments After heating at 1000 C during 1 min (Fig. 11. Sta. Olalla), the purest vermiculite, shows basically enstatite (JCPDS card 19-768); while in the other samples mica and enstatite were formed, the relative proportion of these phases depending on the interlayer composition. Samples of China E and China W, without Mg2+ ions in the interlayer space yielded mica as unique phase; in the other samples (Piau, Gois, Palabora, China G) mica coexists with enstatite. It is the rst time that enstatite is clearly evidenced by powder XRD (Marcos et al., 2004a), Walker and Cole (1957) concluded from thermogravimetric studies that this phase is formed between 810 and 890 C. Suquet et al. (1984) demonstrated by monocrystal XRD studies that enstatite had formed epitaxially on the (010) surface of Mg2+-vermiculite heated at 700 C in air during 24 h. Finally, the SiO2 Al2O3MgO phase diagram (Stoch, 2004) shows that the chemical composition of vermiculite can lead to phases such as enstatite, forsterite, spinel and cordierite. The fact that enstatite appears in the ex situ XRD and it is not observed in the in situ XRD can be explained on the basis of both sample and experiment types: 1) the transformation is faster in powdered samples than in akes (Ruiz-Conde et al.,1996; Marcos et al., 2003, 2004a), and 2) in situ experiments the samples are submitted to a longer gradual and continuous heating (Marcos et al., 2004a) and the experimental conditions are not necessarily in equilibrium and, therefore, kinetics plays a role (Ramrez-Valle et al., 2006).

Fig. 12. SEM image of China E vermiculite (as an example) abruptly heated to 1000 C during 1 min.

Fig. 14. SEM image showing pores with different sizes (pyroexpanded sample of China E, as an example).

C. Marcos et al. / Applied Clay Science 42 (2009) 368378

377

Fig. 15. HRTEM image of wavy and contracted layers (10 ) of Gois vermiculite (as an example) without heating. The circle delimits the zone of differently thick.

3.2.11. SEM study The abrupt desorption of the interlayer water molecules when the samples are heated up to 1000 C during a minute provokes the expansion of the vermiculite particles as revealed by SEM (Fig. 12), leading to a large number of pores of different sizes (Fig. 13), in addition to isolated lamellae of vermiculite other minerals (Fig. 14) could be produced by heating or due to metamorphic reactions during heating. Consequently other factors may also inuence the thermal expansion. This may be the key for obtaining expanded optimized adsorbents for substances of environmental impact. It seems that the expansion after abrupt heating is apparently larger in vermiculites with K+ and/or Na+ and/or Ca2+ with or without Mg2+ in the interlayer space than in the purest vermiculite from Sta. Olalla. These results are in accordance with those of Midgley and Midgley (1960), Couderc and Douillet (1973) and Justo et al. (1989). 3.2.12. Transmission electron microscopy The appearance and the spacing between the layers, close to 10 (Fig. 15), agrees with the patterns described for micas (Buseck and Iijima, 1974). The spacing of 10 corresponds to collapsed vermiculite as a consequence of the dehydration under vacuum (Vali and Kster, 1986; Vali and Hesse, 1992; Marcos et al., 2003. The appearance of 10 -layers demonstrates the beginning of interstratication in addition to structural defects (circled zone in Fig. 16). These observations, also described by Eggleton and Baneld (1985) and Ruz-Cruz (1999) agree with the XRD data and electron microprobe analysis. The undulating form of the layers (Fig. 15), observed also by

Fig. 17. HRTEM image of pyroexpanded vermiculite (Gois, as an example) showing very uniform layers with a spacing of 10 and interstratication with spacings of 19 (area remarked with an ellipse).

other authors (Vali and Hesse, 1992; Ruz-Cruz, 1999), may be attributed to vacuum effects and electrons ow of the TEM. It could be also associated to the distortion of the silicon tetrahedrons that causes a reduction of hexagonal symmetry of the rings into monoclinic. However, the undulation of the layers and the defects observed in HRTEM can also result from interstratication as stated by Marcos et al. (2004b). The location of the lattice fringes of vermiculites from Sta. Olalla, Piau and Gois (with a content of K+ in the interlayer lower than 1) heated abruptly at 1000 C was more difcult than in unheated vermiculites. The reason may be due to the thermal disorder when the new phases coexist with the previous ones with the vermiculite structure. A remarkable difference between the unheated and heated samples is the uniformity of the layers in the heated vermiculites (Fig. 17). The reason could be that without heating the structure collapsed during the observation period, due to the evacuation. The spacing of the layers observed in the samples is close to 10, 16, 17, 19, 21 , although higher values have also been found like in the Palabora samples. The lattice fringes with a spacing of 10 observed in the pyroexpanded samples could be identied as enstatite and/or mica layers. In Sta. Olalla several reasons support the assumption of enstatite: 1) Thermal studies by several authors (Barshad,1948; Suquet et al.,1984; Serratosa and Rausell-Colom, 1985) reported enstatite as the more important phase formed about 800 C in an exothermic reaction. 2) In TEM the spacing d(010) of enstatite shows values of 10 , the same as found in the pyroexpanded sample. 3) Enstatite can grow epitaxially on clay minerals, and 4) the XRD results of the present study also reveals the presence of enstatite (Suquet et al., 1984). The lattice fringes with a spacing 10 in Piau, Gois and Palabora samples may correspond to both mica and enstatite because the XRD patterns at 1000 C obtained by ex situ process reveal the presence of

Table 5 Unit cell parameters, a and b, obtained by electron diffraction a () Sta. Olalla Piau Gois Palabora China W China E 5.49 5.30 4.57 4.81 5.30 4.65 b () 9.36 9.36 8.29 8.45 8.37 8.52

Fig. 16. HRTEM image showing different layers between others pre-existing in the Gois sample, as an example.

378

C. Marcos et al. / Applied Clay Science 42 (2009) 368378 Gonzlez Garca, F., Garca Ramos, G., 1960. On the genesis and transformations of vermiculite. Transactions 7th International Congress of Soil Science, Madison, Wisconsin, vol. 4, pp. 482491. Grim, R.E., 1968. Clay Mineralogy. McGraw-Hill, New York. Hendricks, S.B., Nelson, R.A., Alexander, L.T., 1940. Hydration mechanism of the clay mineral montmorillonite saturated with various cations. Journal of the American Chemical Society 62 (6), 14571464. Hennies, W.T., Stellin Junior, A., 1978. A jazida de vermiculita de Paulistana, Estado do Piaui. Annais do XXX Congresso Brasileiro de Geologia. Recife 4, 17961804. Hindman, J.R., 1992. Vermiculite, Vermiculite Technology, News Letter, 3A, 12p. Justo, A., 1984. Estudio fsico-qumico y mineralgico de vermiculitas de Andaluca y Badajoz. Ph.D. Thesis, Univ. Sevilla, Espaa. Justo, A., Maqueda, C., Prez Rodrguez, J.L., 1986. Estudio qumico de vermiculitas de Andaluca y Badajoz. Boletn Sociedad Espaola de Mineraloga 9, 123129. Justo, A., Maqueda, C., Prez-Rodrguez, J.L., Morillo, E., 1989. Expansibility of some vermiculites. Applied Clay Science 4, 509519. Leonard, R.A., Weed, S.B., 1970. Effects of potassium removal on the b-dimension of phlogopite. Clays and Clay Minerals 18, 197202. Luque, F.J., Rodas, M., Doval, M., 1985. Mineraloga y Gnesis de los yacimientos de vermiculita de Ojen. Boletn de la Sociedad Espaola de Mineraloga 8, 229238. Marcos, C., Argelles, A., Ruz-Conde, A., Snchez-Soto, P.J., Blanco, J.A., 2003. Study of the dehydration process of vermiculites by applying a vacuum pressure: formation of interstratied phases. Mineralogical Magazine 67 (6), 12531268. Marcos, C., Garca, A., Rodrguez Marcos, I., Arango, Y.C., 2004a. Behaviour of vermiculites with different composition at different temperatures. In: Pecchio, M., Andrade, F.R.D., D'Agostino, L.Z., Kahn, H., Sant'Agostino, L.M., Tassinari, M.M.M. (Eds.), Applied Mineralogy: Developments in Science and Technology, ICAM-BR, So Paulo, pp. 627629. Marcos, C., Rodrguez, I., de Renn, L.C., Paredes, J.I., 2004b. Vermiculite surface structure imaged by contact mode AFM. European Journal of Mineralogy 16, 597607. Mathieson, A.M., Walker, G.F., 1954. Crystal structure of magnesium-vermiculite. American Mineralogist 39 (3), 231255. Midgley, H.G., Midgley, C.M., 1960. The mineralogy of some commercial vermiculites. Clay Minerals Bulletin 4 (23), 142150. Mooney, R.W., Keenan, A.G., Wood, L.A., 1952. Adsorption of water vapor by montmorillonite. I. Heat of desorption and application of BET. Journal of the American Chemical Society 74 (6), 13671371. Norris, K., 1973. Factors in the weathering of mica to vermiculite. Proc. Int. Clay. Conf. 1972. Madrd. Divisin de Ciencias, Madrd, pp. 417432. Ramrez-Valle, V., Jimnez-Haro, M.C., Avils, M.A., Prez-Maqueda, L.A., Durn, A., Pascual, J., Prez-Rodrguez, J.L., 2006. Effect of interlayer cations on hightemperature phases of vermiculite. Journal of Thermal Analysis and Calorimetry 84 (1), 147155. Reichenbach, H.G., Beyer, J., 1994. Dehydration and rehydration of vermiculites: IV. Arrangements of interlayer components in the 1.43 nm and 1.38 nm hydrates of Mg-vermiculite. Clay Minerals 29, 327340. Reichenbach, H.G., Beyer, J., 1995. Dehydration and rehydration of vermiculites: II. Phlogopitic Ca-vermiculite. Clay Minerals 30, 273286. Ruiz-Conde, A., Ruiz-Amil, A., Prez-Rodrguez, J.L., Snchez-Soto, P.J., 1996. Dehydrationrehydration in magnesium vermiculite: conversion from twoone and one two water hydration states through the formation of interstratied phases. Journal of Material Chemistry 6, 15571566. Ruiz-Cruz, M.D., 1999. New data for metamorphic vermiculite. European Journal of Mineralogy 11, 533548. Serratosa, J.M., Rausell-Colom, J.A., 1985. Dehydroxylation of Micas and Vermiculites. The Effect of Octahedral Composition and Interlayer Saturating Cations. Mineralogy and Petrography Acta 29, 399408. Shirozu, H., Bailey, S.W., 1966. Crystal structure of a two-layer Mg-vermiculite. American Mineralogist 51, 11241143. Stoch, L., 2004. Thermal analysis and thermochemistry of vitreous into crystalline state transition. Journal of Thermal Analysis and Calorimetry 77, 716. Strand, P.R., Stewart, E., 1983. Vermiculites. In: Lefond, Stanley J. (Ed.), Industrial Mineral and Rocks. The Society of Mining Engineers of the American Institute of Mining, Metallurgical and Petroleum Engineers, Inc., New York, pp. 13751381. Suquet, H., de la Calle, C., Pezerat, H., 1975. Swelling and structural organization of saponite. Clays and Clay Minerals 23 (1), 19. Suquet, H., Mallard, C., Quarton, M., Dubernat, J., Pezerat, H., 1984. Etude du biopyribole form par chauffage des vermiculites magnsiennes. Clay Minerals 19, 217227. Suzuki, M., Wada, N., Hines, D.R., Whittingham, M.S., 1987. Hydration states and phase transitions in vermiculite intercalation compounds. Physical Review. B 36 (5), 28442851. Vali, H., Hesse, R., 1992. Identication of vermiculite by transmission electron microscopy and X-ray diffraction. Clay Minerals 27, 185192. Vali, H., Kster, H.M., 1986. Expanding behaviour, structural disorder, regular and random irregular interstratication of 2:1 layer-silicates studied by high-resolution images of transmission electron microscopy. Clay Minerals 21 (5), 827859. Velasco, F., Casquet, C., Ortega Huerta, M., Rodrguez Gordillo, J., 1981. Indicio de vermiculita en el skarn magntico (aposkarn ogoptico) de La Garrenchosa (Sta. Olalla, Huelva). Boletn de la Sociedad Espaola de Mineraloga 2, 135149. Walker, G.F., 1961. Vermiculite minerals. In: Brown, G. (Ed.), The X-ray Identication and Crystal Structures of Clay Minerals. Mineralogical Society of Great Britain Monograph, pp. 297324. Chap. VII. Walker, G.F., Cole, W.F., 1957. The vermiculite minerals. In: Mackenzie, R.C. (Ed.), The Differential Thermal Investigation of Clays. Mineralogical Society, London.

both minerals. In contrast, the spacing of 10 of the China E and China W vermiculites can be attributed to mica because the XRD patterns at 1000 C show only mica. The unit cell parameters a and b obtained from electron diffraction (Table 5) are higher for the purest vermiculites, i.e. samples of Sta. Olalla and Piau, in agreement with the assumption of Leonard and Weed (1970). 4. Conclusions The commercial vermiculites can be divided into two types: type-1 (Sta. Olalla, Piau and Gois) with Mg2+ and K+ (b 1) as the principal cations in the interlayer space, and type-2 (China E, China W, China G and Palabora) with K+ (= or N 1) and/or Na+ and/or Ca2+ with or without Mg2+ as the principal cations in the interlayer space. The thermal behaviour of both vermiculite types is different. In situ thermal behaviour by heating can be understood considering that the vermiculites constitute a complex system not necessarily in equilibrium and where kinetics plays an important role. The dehydration of type-1 vermiculites seems to be restricted to 1-WLHS, and the dehydroxilation starts to lower temperature and is faster than in type-2 vermiculites. Dehydrated vermiculite of type-2 coexists with a mica-like structure. The maximum hydration state exhibited by the type-2 vermiculites at ambient temperature was equal or lower than 1-WLHS. Enstatite was revealed by the rst time by powder XRD from vermiculite heated abruptly at 1000 C during 1 min. It was the only phase originated in Sta. Olalla vermiculite, in contrast to the other vermiculites which transform to mica-like or mica-like coexisting with enstatite. Vermiculites with higher mica-like content and other phases originated by abruptly heating or due to metamorphic reactions during heating, may be an optimal product for the adsorption of substances of environmental impact, like aromatic compounds, toxic metals or pesticides, due to their higher expansive capacity and higher specic surface. Acknowledgements This study was supported by the Research Project FC-02-PC-CIS0142 from the FICYT (Principado de Asturias, Spain). References
Alcover, J.F., Gatineau, L., 1980. Structure de l'espace interlamellaire des vermiculites Ba monocouches. Clay Minerals 15 (2), 193203. Badreddine, R., 1988. Caractrisation cristalloquimique des vermiculites de Palabora. Rpublique d'Afrique du Sud, et des Bni Bousera, Maroc. Ph.D. Thesis, Univ. Lige, Belgium. Barshad, I., 1948. Vermiculite and its relation to biotite. American Mineralogist 33, 655678. Basset, W.A., 1963. The geology of vermiculite occurrences. Clays and Clay Minerals 10, 6196. Bergaya, F., Theng, B.K.G., Lagaly, G. (Eds.), 2006. Handbook of Clay Science. Elsevier. Brown, G., Brindley, G.W., 1980. X-ray diffraction procedures for clay mineral identication. In: Brindley, G.W., Brown, G. (Eds.), Crystal Structure of Clay Minerals and their X-ray Identication. Mineralogical Society, London, pp. 305359. Buseck, P.R., Iijima, S., 1974. High resolution electron microscopy of silicates. American Mineralogist 59, 121. de la Calle, C., Suquet, H., 1988. Vermiculite. In: Bailey, S.W. (Ed.), Hydrous Phyllosilicates Reviews in Mineralogy, vol. 19. Mineralogical Society of America, Washington, DC, pp. 455496. Collins, D.R., Fitch, A.N., Catlow, R.A., 1992. Dehydration of vermiculites and montmorillonites: a time-resolved powder neutron diffraction study. Journal of Materials Chemistry 8, 865873. Couderc, P., Douillet, Ph., 1973. Les vermiculites industrielles: exfoliation, caractristiques minralogiques et chimiques. Bulletin de la Societe francaise de ceramique 99, 5159. Eggleton, R.A., Baneld, J.F., 1985. The alteration of granitic biotite to chlorite. American Mineralogist 70, 902910. Fster, M.D., 1963. Interpretation of the composition of vermiculites and hydrobiotites. Clays and Clay Minerals 10, 7089. Gevers, T.W., 1949. Vermiculite at Loolekop, Palabora, North East Transvaal. Transactions of the Geological Society of South Africa 51, 133173.

You might also like